Service Science, Vol. 1, No.

1, December 2009 21

THE IMPACTS OF CUSTOMER SATISFACTION ON PROFITABILITY:

A STUDY OF STATE-OWNED ENTERPRISES IN CHINA

Chuan Zhang
School of Economics and Management,
Shanghai Maritime University , China 200135
Xiamen University, China 361005
zhangch@shmtu.edu.cn

Fei Pan
School of Accounting
Shanghai University of Finance and Economics, China 200083
panfei@mail.shufe.edu.cn

ABSTRACT
A growing number of accounting theoreticians believed that non-financial measures can provide incremental information, and
multiple performance measures should be applied to Chinese enterprises. However, little empirical evidence is available on the
behavior of non-financial measures. Using 2-year data from 78 state-owned enterprises, this study provides evidence on the
relationship between non-financial measures and financial performance. The results show that non-financial measure, customer
satisfaction, is significantly associated with contemporary and future financial performance. Furthermore, the enterprises with
higher-level customer satisfaction than others will enjoy higher profitability by increasing unit profit other than enlarging sales.

Keywords: Non-financial measures, Customer satisfaction, Multiple performance measures, Performance measurement.

1. Introduction
Ever since 1980s, there has been much criticism about traditional performance measurement system which is based on financial
measures. Therefore, people have kept exploring appropriate multiple performance measures. Economic theories indicate that,
performance measures should include not only financial measures, but also non-financial measures which reflect different aspects of
the corporate management activities. According to the Agency Theory (Feltham and Xie. 1994; Banker and Datar.1989), if
non-financial measures can provide incremental information of managers‘ behavior apart from financial measures, they should be
used as a basis for compensation decisions. Hence, whether or not the non-financial measures can provide incremental information
has become a common concern among management accountants. An increasing number of articles are discussing about the adoption
of a multiple performance measures system with both non-financial and financial measures (Ittner and Larcker 1998，2001；Amal,
Hassan and Benson 2003；Ittner 1997；Keating,A.,1997). Meanwhile, lots of accounting literature is elaborating on the potential
advantages of non-financial measures (Kaplan and Norton 1992; Kaplan and Atkinson 1989; McNair et al. 1990). Researchers
universally believe that non-financial measures are more future-oriented thus can yield better performance. In an increasing number
of companies, the traditional financial measure has been transformed from the unique performance measurement to a part of
multiple performance measurement system.
Overseas researches on non-financial measures and financial performance are mainly based on customer satisfaction measures,
which attract significant attention from top management in many companies ( K. L. Bernhardt, N. Donthu & P.A. Kennett ,2000).
This sentiment has been echoed in some surveys of upper-level management (Band, 1988; Quinlan, Zemke, Snider, Reinemunde,
Ayling, Singh, Perkins, Antonini, and Loeb, 1991). In a survey of 700 top executives, 64% of them expressed that customer
satisfaction was their first priority, and the rest said it was among their top priorities (Shoultz, 1989).
The majority of these studies have examined the association between customer satisfaction measures and financial or stock
returns. Studies investigating the link between non-financial measures and future financial performance showed mixed results. From
economists‘ perspective, customer satisfaction is a reflection of products and services inputs (Lancaster（1979）and Bowbrick
（1992), but these inputs may not certainly improve a company‘s performance. To achieve high customer satisfaction, a company
always needs heavy investments, which probably lower its profits. For instance, Tornow and Wiley (1991) found a negative
correlation between customer satisfaction and gross profits. The study by Foster and Gupta (1997) of the association between
satisfaction measures for individual customers of a beverage wholesale distributor and its current or future customer profitability
also found positive, negative, or insignificant relations depending upon the questions included in the satisfaction measures or model
specification (levels or percentage changes).
However, most research results actually showed that customer satisfaction is significantly associated with current and future
financial performance. Such researchers as Nelson, Rust, Zahorik, Rose, Batalden, and Siemanski (1992) have found that this
positive relationship exists and it is applied to all profitability measures—earnings, net revenues, and return on assets. Plenty of
 22 Chuan Zhang and Fei Pan

empirical research in the last decade showed that customer satisfaction was positively relevant to corporate performance. Anderson
et al. (1994) studied the relationships between customer satisfaction and the profitability of Swiss companies. They found that
customer satisfaction and ROA (return on assets) are of significantly positive correlation. With the research on customers, operating
entities and companies, Ittner and Larcker（1998）discovered that customer satisfaction and future financial performance are highly
positively correlated, and that there is evidence showing that the publication of customer satisfaction measure will yield incremental
information on stock market. Banker et al.（2000）found the positive correlation between customer satisfaction and financial
performance in 18 hotels run by a company. They also discovered that, when non-financial measures were included in the payment
contracts, managers would attempt to keep in accordance with these non-financial measures and finally improve the corporate
performance.
In the analysis by K. L. Bernhardt, N. Donthu and P.A. Kennett (2000) of a restaurant‘s 342,308 consumer responses, 3,009
employee responses, and its 12-month performance measures, no significant relationship between customer satisfaction and
financial performance was found. But the analysis of time-series data revealed that a positive and significant relationship exists
between changes in customer satisfaction and those in the performance of the company. This may show that the impact of an
increase in customer satisfaction on profits, although obscured in the short run, is significantly positive in the long run.
The research on multiple performance measures in Chinese enterprises are at the stage of experiments and explorations. In
recent years, some Chinese enterprises have introduced non-financial measures to multiple performance measures systems.
However, we still lack empirical evidence to prove whether multiple performance measures are suitable for Chinese enterprises, and
whether non-financial measures can provide incremental information. Viewed from theoretic analyses and foreign researches,
customer satisfaction may bear positive correlation with companies‘ current and future financial performance, but a few scholars
hold that the two may be uncorrelated or even negatively correlated. While in China, there is still room for perfection of market
competition and customers are still unaware of their rights protection, which will to some extent impact the role of customer
satisfaction. Under these conditions, will customer satisfaction still correlated with corporate financial performance? Therefore, this
paper takes 76 state-owned enterprises as research sample and attempts to offer initial answers to this question.

2. Research Background and Data Resource

The performance measurement system in China has its own characteristics compared to those of other countries. It was created
to meet the government‘s needs to manage the state-owned assets rather than to meet the enterprises‘ needs to manage their own
assets. As the owner of SOE, Chinese government developed a series of measures to evaluate managers‘ performance according to
its targets. As the government‘s targets have transformed, the performance measurement systems changed as well.
Before 1970‘s, planned economy is carried out in China. The SOE, as only a subsidiary of the government, had little
self-determination in decision making. The key performance metrics then were production level, quality, profit, cost, safety, and so
on. However, some measures such as profit and cost were not the real indicator of managers‘ efforts and performance because many
decisions were made by the government rather than them.
During 1970‘s and 1980‘s, with the opening and reforms policy in China‘s economy, SOEs gradually became independent
economic organizations. And their managers got more freedom in decision making. Although the old planned economy policy still
had influence, some measures introduced from western countries began to play a positive role in managing state-owned assets.
In 1992, The General Rules Governing Enterprise Financial Affairs (GRGEFA), a primary performance evaluation system, was
issued by Ministry of Finance, in which 8 measures including debt asset ratio, current ratio, quick ratio, account receivable turnover,
inventory turnover, return on equity, profit margin and profit on cost were introduced. According to the GRGEFA, Chinese
government began to evaluate managers‘ performance based on profitability, operation capacity and solvency of state-owned
enterprises they manage. In 1995, evaluation system of enterprise economic benefits, including 10 measures which were divided
into 4 aspects , was issued by the Ministry of Finance. Apart from 3 measures mentioned above, enterprises‘ contribution to society
was considered.
However, these two systems, with only financial measures included, did not work as well in China as expected. There was still
much weakness in management over state-owned assets. Meanwhile, the traditional evaluation system based on financial indicators
has been widely doubted in the western countries, especially with Balanced Scorecard (BSC) of Kaplan and Norton spreading
quickly. Enlightened by BSC, Chinese scholars have been discussing introducing non-financial indicators to Chinese corporate
performance evaluation system since late 1990s.
The past decade has witnessed a remarkable improvement of performance measurement systems in China. in July, 1999, the
Ministry of finance, SETC (the State Economic and Trade Commission), the Ministry of Personnel and SDPC (the State
Development Planning Commission) jointly issued the Rules for State-owned Assets Performance Measurement and the Detailed
Rules for State-owned Assets Performance Measurement (DRSAPM). DRSAPM evaluates state-owned Assets from the perspective
of profitability, operations of state-owned Assets, solvency and growth potential, respectively. On Feb 22, 2002, the Ministry of
finance, SETC(the State Economic and Trade Commission), CWCLE(the Central Work Committee for Large Enterprises), the
Ministry of Labor, the Social Security and Ministry of Labor Protection and SPC(the State Planning Commission) revised the
evaluation rules, and the number of the performance measures have increased to 28, which are categorized into 3 hierarchies–they
 The Impacts of Customer Satisfaction on Profitability 23

are fundamental, amendatory and evaluating measures, which reflect profitability, operations of state-owned Assets, solvency and
growth potential of a state-owned enterprise.
The issuance of Rules for State-owned Assets Performance measurement was a great progress of China‘s performance
measurement system. Different from the traditional unique financial measures systems, the new one included 8 non-financial
measures, such as ―products‘ capacity of earning market share‖.
Based on the revised measurement rules, SASAC-SMG (the State-owned Assets Supervision and Administration Commission
of Shanghai Municipal Government) started the annual measurement of parts of the Shanghai state-owned enterprises, in a way of
measurement by both external assessment companies and enterprises themselves. We thumbed through the performance
measurement reports of 2001 and 2002 and found that, 126 enterprises had presented complete ones in 2001. 55 enterprises in the
126 continued their performance measurement in 2002 because the enterprises who take performance measurement had been chosen
at random, so it is possible for us to get their financial performance of 2002. Apart from those, we got the financial performance of
another 21 enterprises by reviewing their annual financial reports. Totally, the final sample includes 76 enterprises.
According to the definitions of the 10 industry categories of China‘s economy in The Basic Classification of Industries for
Corporate Performance Measurement, the sample covers 8 competitive industries, that is, one enterprise in the catering industry, 9
in the real estate, 30 in the manufacturing , 2 in the construction, 5 in the transportation and logistics, 17 in the wholesale and retail,
3 in the brokerage & agency and 9 in the social services. We believe that only in the competitive industries is ―customer satisfaction‖
correlated with financial performance. Therefore, this sample meets our selection requirements.

3. Research Hypotheses and Approaches

This paper mainly researches on the correlation between a non-financial measure, products‘ capacity of earning market share
(service satisfaction), and financial measures of Chinese enterprises. Overseas researches such as those done by Anderson et al.
(1994) and by Ittner and Larcker (1998), are mainly focused on measuring ‗customer satisfaction‘. As a performance measure,
customer satisfaction can be measured in different ways, like personal interview, telephone interview, questionnaire etc., but the best
way theoretically is from customers. Two main factors determine the accuracy of a customer satisfaction measurement (CSM) study.
The first is asking the right questions; the second is asking the right people - a sample of customers that accurately represents your
customer base (Hill, Brierley and MacDougall, 1999). The sample must be representative, randomly selected and large enough. The
questions might be quite different for different companies. For example, a customer satisfaction survey in the Mass Transit Railway
Corporation of Hong Kong includes the fares, the level of crowdedness, the information about train delays and the air-conditioning
(Megan Walters, 1999). Abdel-Maksouda, Dugdaleb and Lutherc(2005) labeled customer satisfaction as measures of complaints,
returns and warranty claims. And the number of complaints received from customers‘ is classified as a measure of product quality by
Chenhall and Langfield-Smith (1998) but as a measure of customer satisfaction by Drury (1996).
While in the Rules for State-owned Assets Performance measurement, the evaluation measure ―products‘ capacity of earning
market share‖ is divided into two aspects-they are products‘ capacity of earning market share and service satisfaction. In An Notice
on the Distribution of Corporate Performance Measures for the Year 2002 by the General Office of Ministry of Finance released on
March 25, 2002, the best evaluation measures regarding ― products‘ capacity of earning market share‖ are as follows: the qualities
of the prime products measure up to international standards or arrive at foreign advanced standards and certified by ISO
international quality system; leading technical measures come up to the contemporary advanced levels worldwide; the products are
universally accepted by consumers and with reasonable performance versus price ratios; marketing networks are sound and
marketing strategies successful and effective; the products of large enterprises take a lead in the domestic market share in its industry
or nationwide, or attain a 95% production versus sales ratio; the brands enjoy high market reputation and the after-sale services are
excellent, the updating speed is high. While the best measures for ―service satisfaction‖ are as follows: the quality of services is
superior and approved by honorary certificates of national or provincial level in recent 3 years; the service staff have high quality
and amiable and zealous attitude in work, and can meet different demands of customers in time; the products cover a great variety
and services are considerate; the pricing is reasonable and all promises made to customers can be strictly fulfilled. Viewed from the
two aspects of the measure ―products‘ capability of earning market share‖, the major contents of ―customer satisfaction‖ are
included, so we use this measure to replace ―customer satisfaction‖ (CS) in this research.
Unlike general method of measuring ―customer satisfaction‖ which comes from a large sample survey, the measures here
should be graded by more than 5 experts after interviewing managers, analyzing the employee questionnaires and referring to the
standards regulated by the Ministry of Finance. Besides, the experts improved some non-financial measures and made them more
objective in the performance measurement process.
The ―products‘ capacity of earning market share and service satisfaction‖ has been divided into 10 detailed measures (shown in
Figure 1), that is, (1) product quality, focus on the produce ratio up to grade; (2) service quality, focusing on the level of company‘s
service quality and the award they have got in recent 3 years; (3) the guarantee system of service quality and product quality,
focusing on the organization and regulation about the quality management, as well as the implement of them; (4) main technique
standards, focusing on the comparison level among the national and international industry, and if it up to par; (5) the brand, focusing
on the brand strategy of corporation as well as the reputation of their product brand; (6) customer reaction, focusing on the records
 24 Chuan Zhang and Fei Pan

of customer reaction about the service and the times of complaint and praise; (7) marketing strategy, focusing on the strategy and its
efficiency of sale; (8) service diversification, focusing on the reasonable structure of service products; (9) pricing, focusing on the
pricing policy, and if the price is reasonable; (10) deliver the goods and after service, focusing on the regulation of after service and
promise performing.

The quality Service the main the brand

of product quality guarantee technique
system standards

products’ capacity of earning market

share and service satisfaction

customer marketing service pricing after

reaction strategy diversification service

Figure 1. The detail measures of customer satisfaction

In conformity with the research of Anderson et al. (1994), we believe that customer satisfaction of an enterprise is positively
correlated with its current profitability. It can be explained in three aspects: firstly, customer satisfaction not only increase customer
retention rate but could also attract new customers through ―positive word-of-mouth advertising‖ so as to increase sales. Secondly,
customer satisfaction could increase customer loyalty while in turn customer loyalty could decrease transaction expenses and
transaction failure costs. Thirdly, customer satisfaction can be viewed as an asset—that is goodwill of an enterprise. According to
current accounting standards, this kind of self-created goodwill can not be recognized, but it can yield excessive profits assuredly
and thus can promote profitability. We use ROE (return on equity), a generally used financial measure, to reflect the profitability, so
we can suppose here, there is significant positive-correlation between ROE and CS.
H1：There is significant positive-correlation between profitability and CS.
Another problem bubbles up to the surface after we have validated whether CS can influence profitability. That is, if CS can
influence profitability, then in what ways? According to previous literature, it can be explained from the following aspects in the
Figure 2 below:

Increasing
sales

Path 1

Improving CS
Yielding Increasing
goodwill profitability

Path 2

Increasing
profit
margin per
unit

Figure 2. The correlation between CS and profitability

Among the coefficients of sensitiveness concerning profit variance, sales volume is far smaller than unit price and variable costs.
So theoretically, it is easier to improve profitability through increasing profit margin per unit than through expanding sales. To
validate this, we divide ROE into three ratios and then validate the correlation between return on sales and CS, turnover of
shareholders‘ equity and CS, respectively.
 The Impacts of Customer Satisfaction on Profitability 25

netprofit netprofit sales

ROE＝ ＝ 
shareholde rs ' equity sales shareholde rs ' equity
 returnonsa les  turnoverof shareholde rs ' equity

H2：CS increases profitability through expanding sales and increasing profit margin per unit.
In line with the researches of Ittner & Larcker (1998) , We hypotheses that CS is positively correlated with profitability in
following periods. It can be explained from the two aspects: increasing sales and decreasing transaction costs. And there are another
two reasons: firstly, since CS can be viewed as an asset—that is goodwill of an enterprise, goodwill is sure to bring excessive profits
to the enterprise during both current and following periods; secondly, financial measures are usually time-lagged, while
non-financial measures can provide timely information of corporate operation. Non-financial measures in previous period can
usually predict the financial performance in the next period. In the model, we introduce ROE of the year 2001(ROE01) as control
variable and we believe that ROE in 2002 (ROE02) should be positively correlated with ROE01.
H3：CS can predict future profitability of an enterprise.

4. Definition of Control variables

4.1. Measurement institutions (PJJG)

As indicated in the previous parts, SASAC-SMG started a year-after-year measurement of parts of the Shanghai state-owned
enterprises, in a way that measured by both external assessment companies and enterprises group themselves. According to the
regulations of SASAS-SMG, for whether measurement by external assessment companies or self-measurement, the measures should
be graded by more than 5 experts after considering the results of employee questionnaires and referring to the standards regulated by
the Ministry of Finance. So, it is objective to some extent. However, the measures are after all graded by person, are there any
differences in the criterion of different person who does the grading? It is completely possible. One possibility is that the external
assessment companies are even stricter in grading. As assessment companies are professional organization and external institutions,
they may be stricter in measurement. While the managers of the enterprises might just take performance measurement as a
formalistic procedure and because they are familiar with the heads of subsidiary companies, they won‘t give them low grades. The
other possibility is that, the company group is stricter in grading. The leaders of the group might consider from the angle that
measurement would promote management, thus pose strict requirements on subsidiary companies and their heads. While though
assessment companies are external institutions, they are paid by the enterprise itself. So they can not be totally independent because
of the ―charging problem‖. On the contrary, if the group measures its subsidiary companies, this problem wouldn‘t exist. Between
the two possibilities, we are inclined to the second because it involves the problem of economic profits. To sum up, we think that the
choice of measurement institutions will have significant impacts on the grading of this evaluating measure. So, in order to maintain
the objectiveness of research results, we introduce a dummy variable—that is PJJG to represent the choice of measurement
institutions. If the measurement institution is external assessment company, we set it as 1; if parent company, 0.
4.2. Enterprise size （LNASSET）
Same as the vast volume of previous researches, we select the natural logarithm of total assets as the control variable of
enterprise size. Firstly, enterprises of different sizes might have different profitability. As we consider the relative measures of
profitability here, for enterprises with larger size, their ROE might be even lower. Secondly, there might be differences in CS for
enterprises with different sizes. Larger Enterprises may be more powerful in increasing investments to promoting CS. Thirdly, the
size of enterprise may also has connections with the choice of measurement institutions. For instance, the company group might
invite external assessment company to do the measurement of enterprises with larger size, because these enterprises are harder to
measure for its large size. So, we should add control variable of ―enterprise size.‖

5. Descriptive Statistics

We list the definitions of all the variables in Table 1 and do descriptive statistics of them to help people understand the process.
We can see from Table 1 that the mean of ROE01 and that of ROE02 are close to each other, especially for 11.1882 and 11.9351,
only 0.7469 percentage points. The medians of ROE01 and ROE02, 7.83 and 7.25 respectively, are even closer with only 0.58
percentage points variance. It signifies that the economic environment has not changed greatly and so has the average rate of return
in the market. Most of the enterprises have a ROE of 7% to 8%. So in the research we would neglect the macro-economic
environment factors. However, we find that, the range of ROE02 (192.9) is larger than that of ROE01 (116.77), which marks a
worsening polarization in profitability among the 76 enterprises in 2002 against 2001.
Viewed from CS, the highest grade is 16, i.e., full mark, while the lowest is only 9.1, lower than passing mark. It can be
concluded that there are big differences in CS for different enterprises. Then, it needs our further research on whether high CS will
 26 Chuan Zhang and Fei Pan

give rise to better profitability and whether low CS will lead to poorer profitability. In addition, we notice that, the median and mode
of CS are the same, 12.8 (80 percent if according to centesimal system), which means CS for most enterprises is good.
Table 1. The definitions of variables and descriptive statistics
CS PJJG LNASSET ROE01 ROS01 JZCZZL ROE02
Definition of natural Return on Return on turnover of Return on
Customer Measurement
variables logarithm of equity in 2001 sales in 2001 shareholders‘ equity in
satisfaction institution
total assets （％） （％） equity 2002（％）
Mean 13.3355 0.3289 9.9235 11.1882 0.0822 4.0765 11.9351
Standard
0.20354 0.05425 0.22535 1.93076 0.02091 1.13117 2.51977
error
Median 12.8000 0.0000 9.9383 7.8300 0.0422 1.5526 7.2500
Mode 12.80 0.00 5.77(a) -22.67(a) -0.66(a) 0.02(a) 1.03(a)
Variance 3.149 0.224 3.860 283.314 0.033 97.246 482.543
Range 6.90 1.00 8.50 116.77 1.52 76.87 192.90
Maximum 9.10 0.00 5.77 -22.67 -0.66 0.02 -28.00
Minimum 16.00 1.00 14.27 94.10 0.85 76.89 164.90
(a) There exist many modes and the listed one is the smallest of them.

6. Analyses Results and Discussions

6.1. Comparison of means in the sampling groups

Based on the statistics of the marks of CS, we find that 18 enterprises get a score of 12.8, 21 get a score lower than 12.8 while
37 get a score higher than 12.8. As shown in Figure 3.
20

15
Count

10

9.10 10.20 11.50 12.80 13.40 14.70 15.40

9.60 10.90 12.20 13.30 14.10 14.90 16.00
CS

Figure 3.Statistics of the marks of CS

In order to directly reflect the correlation between CS and financial performance, we define those enterprises with a score lower
than 12.8 as L-CS sampling group and those with a score higher than 12.8 as H-CS sampling group. And then we compare the means
of ROE01，ROS01， ROE02， JZCZZL in these two groups and do T-test on the means. As shown in Table 2.

Table 2. Comparison of means in two sampling groups

CS level Mean Variance of means t The value of P
ROE01 H-CS 16.7868
12.4791 3.275*** 0.002
L-CS 4.3076
ROS01 H-CS 0.1222
0.11197 2.171** 0.034
L-CS 0.0102
ROE02 H-CS 18.7527
15.3757 2.413** 0.019
L-CS 3.3770
JZCZZL H-CS 2.3470
-4.28590 -1.166 0.257
L-CS 6.6329
 The Impacts of Customer Satisfaction on Profitability 27

We find that, except that means of JZCZZL in the two sampling groups are not significantly different, means of ROE01,
ROS01, ROE02 in the two groups are significantly and positively correlated. In other words, the means of ROE01, ROS01, ROE02
are higher in the group with higher CS than that with lower CS. It proves to some extent that CS impacts ROE and CS impacts ROE
through impacting ROS rather than the turnover of shareholders‘ equity. It can be put in another way that, CS improves corporate
profitability through increasing profit margin per unit other than expanding sales. Here we preliminarily validate H1, H2 and H3.
6.2. Analyses of correlation coefficients
To further validate the previous hypotheses, we analyze the correlation coefficients of all the variables and do the two-tailed
tests. The results are shown in Table 3.

Table 3. Analyses of correlation coefficients of all the variables

PJJG ROE01 ROS01 JZCZZL LNASSET ROE02
CS 0.443*** 0.290** 0.265** -0.138 0.409*** 0.322***
0.000 0.011 0.021 0.235 0.000 0.005
PJJG 1 0.208* -0.024 -0.057 0.253** 0.058
. 0.071 0.835 0.623 0.027 0.617
ROE01 1 0.334*** -0.016 -0.035 0.803***
. 0.003 0.889 0.767 0.000
ROS01 1 -0.122 -0.083 0.487***
. 0.295 0.475 0.000
JZCZZL 1 -0.185 -0.015
. 0.109 0.897
LNASSET 1 -0.067
. 0.564
*** Correlation is significant at the 0.01 level (2-tailed).
** Correlation is significant at the 0.05 level (2-tailed).
* Correlation is significant at the 0.1 level (2-tailed).
From the analyses of correlation coefficients of all the variables we could see that, CS has significant positive correlation with
the choice of measurement institutions, current-year ROE, current-year ROS, natural logarithm of total assets and the following-year
ROE, which is in line with H1, H2 and H3. There is significant positive correlation at the level of 0.01 between CS and the choice
of measurement institutions, which indicates that the CS of enterprises who choose an external assessment company to do the
measurement is higher than that of those who choose to do self-measurement. CS is correlated with current-year ROS but not
significantly correlated with JZCZZL, which signifies that the improvement of profitability (the increase of ROE) is mainly because
the increase of CS levels up profit margin per unit and finally improves profitability rather than improving profitability via
expanding sales. It is in conformity with H2. CS is significantly and positively correlated with natural logarithm of total assets at the
level of 0.01, which represents that enterprises with larger size are usually more powerful and can input more to improve CS, and
besides, they enjoy high and good reputation in the industry, which is what we defined as goodwill. So it is no doubt that they will get
higher score of CS than other enterprises.
In addition, we notice that there is also significant positive correlation between current-year ROE and the choice of
measurement institutions, natural logarithm of total assets and the choice of measurement institutions, respectively. On one hand, it
indicates that the company group is inclined to invite external assess companies to do measurement on the subsidiary companies
with good financial performance while leave those with relatively poor financial performance for self-measurement. That is what we
called ―Do not wash your dirty linen in public‖. On the other hand, enterprises with larger size usually have more sophisticated
organizations and businesses so it is more difficult to measure performance. While external assess company are professional
measurement institutions so enterprises with large size are inclined to invite external assess companies to do the measurement.
6.3. C. Regression analyses
Based on H1, H2and H3, we form the following regression models. The regression results are shown in Table 4.
ROE01   0  1CS   2 PJJG   3 LNASSET Model－1
ROS 01   0   1CS   2 PJJG   3 LNASSET Model－2a
JZCZZL   0  1CS   2 PJJG   3 LNASSET Model－2b
ROE02   0   1CS   2 PJJG   3 ROE01   4 LNASSET Model－3
From Table 4 we could see that, the regression results are completely in line with H1 and H2. CS is significantly and positively
correlated with both current-year ROE and current-year ROS but not correlated with turnover of shareholders‘ equity. Besides, the
regression models Model－1 and Model－2a both pass F-test and do not exist the problem of multiple linearity by view from the
value of VIF,.
 28 Chuan Zhang and Fei Pan

Table 4: regression analyses and the results of T-test

Dependent variables CS PJJG LNASSET F-Value
β 3.019 4.126 -1.663
t 2.429** 0.938 -1.599
ROE01 Sig. 3.372**
0.018 0.351 0.114
VIF 1.41 1.21 1.254
β 0.043 -0.061 -0.020
t 3.248*** -1.278 -1.795*
ROS01 Sig. 3.708**
0.002 0.205 0.077
VIF 1.410 1.254 1.210
β -0.456 0.392 -0.786
t -0.598 0.145 -1.233
JZCZZL 0.981
Sig. 0.551 0.885 0.222
VIF 1.410 1.254 1.210
β 2.562 -8.106 -0.904
t 2.516** -2.325** -1.084
ROE02 38.601***
Sig. 0.014 0.023 0.282
VIF 1.526 1.270 1.253
*** Correlation is significant at the 0.01 level (2-tailed).
** Correlation is significant at the 0.05 level (2-tailed).
* Correlation is significant at the 0.1 level (2-tailed).
The result of Model－3 tells us that CS bears significant positive correlation with the second-year ROE. When analyzing the
correlation coefficients in the previous parts, we discover that company group is inclined to invite external assessment companies to
do the measurement on the subsidiary companies with higher ROE01. While in the regression of ROE02 based on the control
variable ROE01, we find negative correlation, which indicates that the company group, who does self-measurement, has to some
extent realized the purpose of improving management via measurement. Because the leaders of groups participate in the
measurement in person, the problems in management found in the measurement will be solved and in the work of next period, the
management will be improved. So the changes of profitability for them are better than those for the enterprises who invite external
assessment companies to do the measurement. In addition, both of these two regression model have passed F-test and viewed from
the value of VIF there do not exist the problem of multiple linearity.

7. Conclusion
The research confirms that for Chinese SOEs non-financial measures of CS is positively correlated with financial performance.
It also shows that those SOEs with higher CS scores enjoy higher profitability. Additionally, SOEs focusing on CS are more likely
to improve profitability through increasing unit profit margins rather than by merely expanding sales. It seems that the use of CS
measures not only predicts future financial performance but also its changes.
These findings confirm some of the results of the previous research however what is unique in the context of this research and
China is that no direct reference has been made to customers although the Chinese model takes a wider approach in interpreting CS.
Besides regression used by many researchers, compare mean was used in this study.
Although some literatures have identified the positive association between non-financial measures and financial performance,
such as Nelson, Rust, Zahorik, Rose, Batalden, and Siemanski (1992), Anderson, Fornell and Lehmann's（1994）, Ittner and
Larcker(1998), Yeung and Ennew (2000) , Said, Hassan, Elnaby and Wier, (2003), Miguel et al. (2004), this study makes a unique
contribution. Different from prior researches generally based on surveys, this study collected its data from the performance reports
of Chinese SOEs controlled by the State-owned Assets Supervision and Administration Committee of Shanghai. First, these data are
very reliable because we got them from the government where the reports are handed in. Second, there are less noises because the 76
enterprises have same corporate governance. Unlike general methods of measuring ―customer satisfaction‖ which comes from a
survey or National American Customer Satisfaction Index (ACSI), the variable ―customer satisfaction‖ here was measured by
trained experts after interviewing managers, analyzing the employee questionnaires and referring to the standards regulated by the
Ministry of Finance.
Considering two types of measuring institutions involved in performance measurement, a dummy variable PJJG was
introduced to models. The results indicate that CS is positively related to the nature of expert organizations, with external experts
scoring CS more highly than internal experts. Another control variable, enterprise size, was introduced in the models and indicates
 The Impacts of Customer Satisfaction on Profitability 29

that the size of the target organizations is also positively correlated to CS.
After the positive relationship between Customer Satisfaction and profitability is being found, a further research has been
conducted to understand how the Customer Satisfaction influences the profitability. Without considering unit profit margins, Ittner
& Larcker（1998）only found the positive relationship between sales and Customer Satisfaction. However, according to management
accounting theory, the change of price and direct cost drives profit more sensitively. Therefore, theoretically, compared with the
increasing sales, the increasing unit profit margin can play a bigger role in improving profitability . If we split return on equity (ROE)
to return on sales (ROS) and turnover of shareholder equity, we can find a significant relationship between customer satisfaction and
ROS instead of a significant relationship between customer satisfaction and turnover of shareholder Equity. Although our findings
are inconsistent with the findings of Ittner and Larcker, it is supported by the theories of management accounting. Results indicate
that corporate profitability is increased via improved unit profit margins rather than increased sales.
This research confirms the significant impact of customer satisfaction on profitability in Chinese State-owned Enterprises, It is
an evidence of the necessity of constructing a multiple performance measure system in China.

Acknowledgement
The authors are grateful for funding (No.70671063) provided by Chinese Natural Science Research Institute and Shanghai City
Pujiang Program (No: PJ200800551). We also would like to thank anonymous reviewers for their constructive comments on the
earlier version of the manuscript, as well as the research assistance provided by Ying Wang and Shuai Jing. The data in this paper
were collected from the enterprises‘ performance evaluation report of the Shanghai State-owned Assets Supervision and
Administration Commission of Shanghai Municipal Government, to which we want to show our acknowledgement.

REFERENCES
Ahmed Abdel-Maksouda, David Dugdaleb and Robert Lutherc, ―Non-financial performance measurement in manufacturing
companies,‖ The British Accounting Review 37, 261–297, 2005.
Amal A. Said, Hassan R. Hassab Elnaby, and Benson Wier, ―An empirical investigation of the of the performance consequences of
nonfinancial measures,‖ Journal of management Accounting Research 15, 193-223, 2003.
Anderson, E., C. Fornell, and D. Lehmann. ―Customer satisfaction, market share, and profitability: Findings from Sweden,‖ Journal
of Marketing (July): 53-66, 1994.
Band, William, ―Measuring the High Payoff From Satisfied Customers,‖ Sales & Marketing Management in Canada, 23-25,
December 1988.
Banker, R.D., and S. M. Datar, ―Sensitivity, precision, and linear aggregation of signals,‖ Journal of Accounting Research 27(1),
21-40, 1989.
Banker, R.D, Potter, G., and Srinivasan D., ―An empirical investigation of an incentive plan that includes non-financial performance
measures‖. The Accounting Review 75, 65-92, 2000.
Chenhall, R.H. and Langfield-Smith, K., ―Adoption and benefits of management accounting practices: an Australian study,‖
Management Accounting Research 9, 1–19, 1998.
Chinese Ministry of Finance, General Rules Governing Enterprise Financial Affairs, Approved by the State Council on November
16, 1992 and promulgated by Decree No. 4 of the Ministry of Finance on November 30, 1992.
Drury, C., Management and Cost Accounting, fourth edition. International Thomas Business Press, London, 1996.
K. L. Bernhardt, N. Donthu & P.A. Kennett, ―A Longitudinal Analysis of Satisfaction and Profitability,‖ Journal of Business
Research 47, 161-171, 2000.
Feltham, G. A. and J. Xie, ―Performance measures congruity and diversity in multi-task principal / agent relations,‖ The Accounting
Review 69(July):429-453,1994.
Ittner C. D. and David F. Larcker, ―Are non-financial measures leading indicators of financial performance? An analyses of
customer satisfaction,‖ Journal of Accounting Research (supplement):1-35, 1998.
Ittner C. D. and David F. Larcker, ―Innovations in performance measurement: trends and research implications,‖ Journal of
management Accounting Research 10, 205-238, 1998.
Kaplan. S. and D. Norton. ―The balanced-scorecard: Measures that drive performance,‖ Harvard Business Review
(January-February):71-79, 1992.
Kaplan. S. and A. Atkinson, Advanced Management Accounting, Second edition, Englewood Cliffs, NJ: Prentice Hall,1989.
Keating, A., ―Determinants of divisional performance evaluation practices,‖ Journal of Accounting and Economics 24, 243-273,
1997.
McNair, Carol J., Lynch, Richaed L., Cross, Kelvin F., ―Do financial and non-financial performance measures have to agree?‖
Management Accounting Nov.28-36.1990
Megan Walters, ―Performance measurement systems –a case study of customer satisfaction,‖ Facilities, Volume 17 Number 3/4
97–104, March/April 1999
Nelson, Eugene C., Rust, Roland T., Zahorik, Anthony, Rose, Robin L, Batalden, Paul, and Siemanski, Beth Ann, ―Do Patient
Perceptions of Quality Relate to Hospital Financial Performance?‖ Journal of Health Care Marketing, 6–13, December 1992.
Nigel Hill, John Brierley and Rob MacDougall, How to Measure Customer Satisfaction, Gower publishing Limited, 1999.
 30 Chuan Zhang and Fei Pan

Quinlan, Michael R., Zemke, Ron, Snider, Jim, Reinemund, Steven S., Ayling, Robert, Singh, Karmjit, Perkins, James A., Antonini,
Joseph E., and Loeb, Walter F., ―How Does Service Drive the Service Company?‖ Harvard Business Review
(November–December), 146–158, 1991.
Shoultz, Donald, ―Service Firms Are Customer Driven: Study,‖ American Banker (February 23), 17–18, 1989.
Tornow, Walter W., and Wiley, Jack W., ―Service Quality and Management Practices: A Look at Employee Attitudes, Customer
Satisfaction, and Bottom-Line Consequences,‖ Human Resource Planning 14, 105–115, 1991.